In spite of the huge progress in the treatment of diabetes mellitus, we are still in the situation that both pregestational (PGDM) and gestational diabetes (GDM) impose an additional risk to the embryo, fetus, and course of pregnancy. PGDM may increase the rate of congenital malformations, especially cardiac, nervous system, musculoskeletal system, and limbs. PGDM may interfere with fetal growth, often causing macrosomia, but in the presence of severe maternal complications, especially nephropathy, it may inhibit fetal growth. PGDM may also induce a variety of perinatal complications such as stillbirth and perinatal death, cardiomyopathy, respiratory morbidity, and perinatal asphyxia. GDM that generally develops in the second half of pregnancy induces similar but generally less severe complications. Their severity is higher with earlier onset of GDM and inversely correlated with the degree of glycemic control. Early initiation of GDM might even cause some increase in the rate of congenital malformations. Both PGDM and GDM may cause various motor and behavioral neurodevelopmental problems, including an increased incidence of attention deficit hyperactivity disorder (ADHD) and autism spectrum disorder (ASD). Most complications are reduced in incidence and severity with the improvement in diabetic control. Mechanisms of diabetic-induced damage in pregnancy are related to maternal and fetal hyperglycemia, enhanced oxidative stress, epigenetic changes, and other, less defined, pathogenic mechanisms.

Since the first animal model of valproic acid (VPA) induced autistic-like behavior, many genetic and non-genetic experimental animal models for Autism Spectrum Disorder (ASD) have been described. The more common non-genetic animal models induce ASD in rats and mice by infection/inflammation or the prenatal or early postnatal administration of VPA. Through the establishment of these models, attempts have been made to ameliorate or even prevent ASD-like symptoms. Some of the genetic models have been successfully treated by genetic manipulations or the manipulation of neurotransmission. Different antioxidants have been used (i.e., astaxanthin, green tea, piperine) to reduce brain oxidative stress in VPA-induced ASD models. Agents affecting brain neurotransmitters (donepezil, agmatine, agomelatine, memantine, oxytocin) also successfully reduced ASD-like symptoms. However, complete prevention of the development of symptoms was achieved only rarely. In our recent study, we treated mouse offspring exposed on postnatal day four to VPA with S-adenosine methionine (SAM) for three days, and prevented ASD-like behavior, brain oxidative stress, and the changes in gene expression induced by VPA. In this review, we describe, in addition to our data, the existing literature on the prevention/amelioration of ASD-like symptoms. We also discuss the possible mechanisms underlying some of these phenomena. Finally, we describe some of the clinical trials in children with ASD that were carried out as a result of data from animal studies, especially those with polyunsaturated fatty acids (PUFAs).

S-adenosylmethionine (SAMe) is involved in many transmethylation reactions in most living organisms and is also required in the synthesis of several substances such as monoamine neurotransmitters and the N-methyl-D-aspartate (NMDA) receptor. Due to its important role as an epigenetic modulator, we discuss in some length the process of DNA methylation and demethylation and the critical periods of epigenetic modifications in the embryo, fetus, and thereafter. We also discuss the effects of SAMe deficiency and the attempts to use SAMe for therapeutic purposes such as the treatment of major depressive disorder, Alzheimer disease, and other neuropsychiatric disorders. SAMe is an approved food additive and as such is also used during pregnancy. Yet, there seems to scanty data on the possible effects of SAMe on the developing embryo and fetus. Valproic acid (VPA) is a well-tolerated and effective antiepileptic drug that is also used as a mood stabilizer. Due to its high teratogenicity, it is contraindicated in pregnancy. A major mechanism of its action is histone deacetylase inhibition, and therefore, it acts as an epigenetic modulator, mainly on the brain. This prompted clinical trials using VPA for additional indications i.e., treating degenerative brain disease such as Alzheimer disease, dementia, HIV, and even cancer. Therefore, we discuss the possible effects of VPA and SAMe on the conceptus and early postnatally, during periods of susceptibility to epigenetic modifications. VPA is also used as an inducer of autistic-like behavior in rodents and was found by us to modify gene expression when administered during the first postnatal week but not when administered to the pregnant dams on day 12 of gestation. In contrast, SAMe modified gene expression when administered on day 12 of pregnancy but not postnatally. If administered together, VPA prevented the changes in gene expression induced by prenatal SAMe administration, and SAMe prevented the gene expression changes and autistic-like behavior induced by early postnatal VPA. It is concluded that both VPA and SAMe are powerful epigenetic modifiers with antagonistic actions on the brain that will probably be used in the future more extensively for the treatment of a variety of epigenetic diseases of the nervous system.

In previous studies we produced autism like behavioral changes in mice by Valproic acid (VPA) with significant differences between genders. S-adenosine methionine (SAM) prevented the autism like behavior in both genders. The expression of 770 genes of pathways involved in neurophysiology and neuropathology was studied in the prefrontal cortex of 60 days old male and female mice using the NanoString nCounter. In females, VPA induced statistically significant changes in the expression of 146 genes; 71 genes were upregulated and 75 downregulated. In males, VPA changed the expression of only 19 genes, 16 were upregulated and 3 downregulated. Eight genes were similarly changed in both genders. When considering only the genes that were changed by at least 50%, VPA changed the expression of 15 genes in females and 3 in males. Only Nts was similarly downregulated in both genders. SAM normalized the expression of most changed genes in both genders. We presume that genes that are involved in autism like behavior in our model were similarly changed in both genders and corrected by SAM. The behavioral and other differences between genders may be related to genes that were differently affected by VPA in males and females and/or differently affected by SAM.

The etiology of autism spectrum disorder (ASD) is genetic, environmental, and epigenetic. In addition to sex differences in the prevalence of ASD, which is 3–4 times more common in males, there are also distinct clinical, molecular, electrophysiological, and pathophysiological differences between sexes. In human, males with ASD have more externalizing problems (i.e., attention-deficit hyperactivity disorder), more severe communication and social problems, as well as repetitive movements. Females with ASD generally exhibit fewer severe communication problems, less repetitive and stereotyped behavior, but more internalizing problems, such as depression and anxiety. Females need a higher load of genetic changes related to ASD compared to males. There are also sex differences in brain structure, connectivity, and electrophysiology. Genetic or non-genetic experimental animal models of ASD-like behavior, when studied for sex differences, showed some neurobehavioral and electrophysiological differences between male and female animals depending on the specific model. We previously carried out studies on behavioral and molecular differences between male and female mice treated with valproic acid, either prenatally or early postnatally, that exhibited ASD-like behavior and found distinct differences between the sexes, the female mice performing better on tests measuring social interaction and undergoing changes in the expression of more genes in the brain compared to males. Interestingly, co-administration of S-adenosylmethionine alleviated the ASD-like behavioral symptoms and the gene-expression changes to the same extent in both sexes. The mechanisms underlying the sex differences are not yet fully understood.

In this review, we discuss the functions and main effects on pregnancy outcomes of three agents that have the ability to induce epigenetic modifications: valproic acid (VPA), a well-known teratogen that is a histone deacetylase inhibitor; S-adenosylmethionine (SAMe), the most effective methyl donor; and choline, an important micronutrient involved in the one methyl group cycle and in the synthesis of SAMe. Our aim was to describe the possible effects of these compounds when administered during pregnancy on the developing embryo and fetus or, if administered postnatally, their effects on the developing child. These substances are able to modify gene expression and possibly alleviate neurobehavioral changes in disturbances that have epigenetic origins, such as autism spectrum disorder (ASD), depression, Rett syndrome, and fetal alcohol spectrum disorder (FASD). Valproic acid and SAMe are antagonistic epigenetic modulators whether administered in utero or postnatally. However, VPA is a major human teratogen and, whenever possible, should not be used by pregnant women. Most currently relevant data come from experimental animal studies that aimed to explore the possibility of using these substances as epigenetic modifiers and possible therapeutic agents. In experimental animals, each of these substances was able to alleviate the severity of several well-known diseases by inducing changes in the expression of affected genes or by other yet unknown mechanisms. We believe that additional studies are needed to further explore the possibility of using these substances, and similar compounds, for the treatment of ”epigenetic human diseases”.